246 THE INTELLIGENCE OF MAMMALS 
the door is shut— he makes off round two corners to the 
front door, and so into the dining-room. He had never 
been in this room before, but has once been from the back 
into the house by the front door. The experiment is once 
repeated, and the dog remembers this route five days later. 
On arriving at the house on this occasion, he is taken through 
a side door into the dining-room, and then out at the back. 
He first finds his way in through the front as mentioned, and 
then for a further trial both front and back door are shut. 
The dog goes to and fro from one door to the other, and then 
suddenly goes right off around the house, and in by the side 
door— a route which he had never taken before. There may 
have been an element of chance in this success, but, on the 
whole we seem to have a series of acts dictated by the desire 
to find the master operating on the remembrance of the 
modes of entrance.” 
The ability of animals such as horses to find their way 
back for miles over a road which they have only followed 
once is indicative of something more than mere sensori- 
motor association. I well remember a horse we once 
owned whose memory for the proper turns in the road he 
had taken in going away from home I had often tested and 
found to be almost infallible. Like many other horses he 
was a much more willing traveller when homeward bound, 
but whether he was influenced by an idea of hay and oats 
and rest to be enjoyed at the end of his journey it might be 
hazardous to say. If the homing of the animal were due 
to a blind sensori-motor association, we should have to 
assume that the sight of particular objects along his course 
came to be associated with particular movements; object 
A, for instance, with a slight turn to the right, and object 
B with a slight turn to the left, and so on. If the animal 
passed over the road again in the same direction we might 
